An engineering drawing is a type of technical drawing which may be used to define requirements for engineered items. The engineering drawing typically contains engineering diagrams and associated text components. The text components may contain important information in textual form around the engineering diagrams. The text components may be tag numbers, serial numbers, names and so on. Separating the text components from the engineering drawings is a challenging task, and such separation may differ in many aspects from non-engineering drawing documents. OCR technology is mainly used to convert virtually any kind of images containing written text into machine-readable text data. However, a system, without help of an OCR system, may not be able read the engineering drawings which are available in form of scanned images. Extraction of the textual components around the graphical symbol and non-text elements becomes difficult as filtration of only text is required before passing to the OCR system. Poor result in the OCR system may be achieved if the engineering drawings are directly passed to the OCR system, without removal of graphics or drawings. The poor performance of the OCR system on the engineering drawings may be due to lack of prior knowledge about separation of the text components and the drawing components. The OCR system may perform poorly even if the probable text candidates are marked.
One or more technologies are developed to improve the OCR system. Some technologies include extraction of the text components from the engineering drawings. Such technologies include creating a binary representation of a drawing to form an electronic image of pixels and comparing pixel values to a gray-scale threshold value and assigning the pixels to be either white or black. The image is discriminated between text regions and graphical lines in the image by determining a distance between each pixel and nearest neighboring pixels. The distance is compared with a predetermined distance and labels are assigned to the pixels which are within the predetermined distance, to group pixels into blocks. Blocks with such pixels are compared with a predetermined format to identify possible text regions. The graphical lines that remain in the text regions are removed to create text-only regions. The text is recognized within the text only regions by using OCR system.
Other technologies in the OCR system include a new rule-based method for text and graphics separation based on analysis of text and graphics features of general engineering drawings. Such approaches propose to erase non-text regions from engineering drawings as much as possible, rather than extracting text regions. Some technologies include machine learning algorithms to detect the components to be character or not.
However, such technologies include limitations on kind of engineering drawings and noise level. Also, such technologies may not be robust to the engineering drawings with varying fonts and sizes. For the engineering drawings where distance between the text components and the drawing components is minimal or zero, such technologies may provide an inaccurate output.
The information disclosed in this background of the disclosure section is only for enhancement of understanding of the general background of the invention and should not be taken as an acknowledgement or any form of suggestion that this information forms the prior art already known to a person skilled in the art.